Intelligent Systems, Robotics, and Control
Prof. Kenneth Kreutz-Delgado
Jacobs Hall, 6608
This information sciences-based field is concerned with the design of human-interactive intelligent systems that can sense the world (defined as some specified domain of interest); represent or model the world; detect and identify states and events in the world; reason about and make decisions about the world; and/or act on the world, perhaps all in real-time. A sense of the type of systems and applications encountered in this discipline can be obtained by viewing the projects shown at the website http://vision.ucsd.edu/.
The development of such sophisticated systems is necessarily an interdisciplinary activity. To sense and succinctly represent events in the world requires knowledge of signal processing, computer vision, information theory, coding theory, and data-basing; to detect and reason about states of the world utilizes concepts from statistical detection theory, hypothesis testing, pattern recognition, time series analysis, and artificial intelligence; to make good decisions about highly complex systems requires knowledge of traditional mathematical optimization theory and contemporary near-optimal approaches such as evolutionary computation; and to act upon the world requires familiarity with concepts of control theory and robotics. Very often learning and adaptation are required as either critical aspects of the world are poorly known at the outset, and must be refined on-line, or the world is non-stationary and our system must constantly adapt to it as it evolves. In addition to the theoretical information and computer science aspects, many important hardware and software issues must be addressed in order to obtain an effective fusion of a complicated suite of sensors, computers, and problem dynamics into one integrated system.
Faculty affiliated with the ISRC subarea are involved in virtually all aspects of the field, including applications to intelligent communications systems; advanced human-computer interfacing; statistical signal- and image-processing; intelligent tracking and guidance systems; biomedical system identification and control; and control of teleoperated and autonomous multiagent robotic systems.